URNADES

Universal RNA Differential Expression Scripts

This R scripts perform differential expression analysis (DE) on RNA-seq data. They can be used with either featureCounts or SALMON data-folder.

For example HTML reports please see the example directory.

Usage

To run, you will need to provide the following information:

The path to the sample information file (a .csv with at least sampleID and Treatment columns).
The formula for the DE analysis (batch example: "~ Treatment", or paired sample: "~ PatientID + Treatment").
The name of the condition of interest (example: "Treatment").
The path to the featureCounts or SALMON data-folder.
The output path.
The minimum count (optional, default = 10).
The FDR value (optional, default = 0.05).
The minimal Log2 fold change.

The featureCounts should be combined into a counts.tsv file (column per sample, row per gene). Path to the SALMON files can be given directly -- the parent directory, which contains sampleID/quant.sf. Make sure the sampleID matches the one in sample_info.csv.

References

If left unspecified, human references will be downloaded from Ensembl: biomaRt::useMart(biomart = "ensembl", dataset = "hsapiens_gene_ensembl")

Otherwise use the following parameters to specify custom annotations:

--annotatedPath - Path to annotation csv [gene_id, gene_symbol, description], gtf or gff
--t2gPath - Path to genes.filtered.t2t

Environment

To ensure the presence of all required R libraries simply crete a conda env from the environment.yaml:

conda env create -f environment.yaml
-or-
mamba env create -f environment.yaml

conda activate urnades

You can also install R packages locally with the help of renv. Make sure that the renv.lock file is present.

# Open R terminal in project directory
# Make sure the R library renv is installed locally

renv::restore()

For example, to run the script with featureCounts & SALMON data-folder, you would use the following command:

Rscript URNADES.R \
--sampleInfoFilePath sample_info.csv \
--featureCountsdata $featureCounts_data_path \
--SALMONdata $SALMON_data_path \
--output $output \
--formula_input "~ PatientID + Treatment" \
--conditionName "Treatment" \

The sample_info.csv should look like this:

SampleID,PatientID,Treatment
C1_S3,Patient_1,Pre
D1_S4,Patient_1,Semaglutid
G1_S7,Patient_2,Pre
H1_S8,Patient_2,Semaglutid

Output

The scripts will create two HTML reports in the output directory:

• featureCounts_report.html: This report contains the results of the DE analysis for featureCounts data.
• SALMON_report.html: This report contains the results of the DE analysis for SALMON data-folder.

Additionally, a subfolder containing analysis result data tables is added.

License

This scripts are licensed under the MIT License.